Private sector sugar millers and Ithala Development Finance Corporation (Ithala) implemented a graduated mortgage loan repayment scheme in the 1995/96 sugarcane production-season to try and improve access to farmland by aspirant commercial farmers in KwaZulu-Natal. By March 2001, the scheme had financed 106 "medium scale farmers" (MSFs), 99 of whom were still in the scheme (one loan had been repaid from own funds, and another six from the proceeds of life insurance policies). The first aim of this study was to analyse factors affecting whether or not the MSFs were current or in arrears on loan repayments as at 31 March 2001. A logit model based on full information for 83 MSFs shows that the estimated probability of a MSF being current on loan repayments was higher for clients with higher levels of average annual gross turnover relative to loan size, and for clients with access to substantive off-farm income. This suggests that farm size (proxied by annual farm gross turnover) does matter when policymakers in South Africa consider future similar schemes designed to improve access to commercial farmland by people that previously could not buy farmland. Smaller-sized, creditworthy farms with loan sizes that are relatively low compared to the expected average annual gross income may also be viable. Access to off-farm income could also be considered as a criterion in selecting potential farmers for future similar schemes, as it helps to provide additional liquidity to fund future operations and debt repayments, and can reduce leverage levels. The second aim was to conduct personal interviews with the 99 MSFs between July and September 2001 in order to identify what aspects of the scheme could be improved for new members . Responses from 88 of these MSFs show that 68% of them would opt to first rent land before purchasing, while 78% of them recognize, or have experienced, the cash flow problem associated with land purchase. Most of the MSFs felt that long-term sugarcane supply agreements constrain enterprise diversification, and that the quality of mentorship that they currently received was not satisfactory. Industry players could consider leveraging donor funding for empowerment projects to improve the quality of future mentorship programmes. There is also some scope for Ithala to improve the client-lender relationship by better clarifying the structure of the graduated repayments, sending loan statements on time, and helping clients to interpret loan statements. Growers perceive the need for a coordinator to monitor, and advise on how to improve, their financial performance this could be a new commercial service opportunity. Using an independent valuer to conduct farm valuations may also be necessary to avoid perceptions of bias in the value of farms offered for sale by the millers. A logit model of the MSFs' preferences for first renting land before purchase shows that new growers joining this scheme, or similar schemes for other farm products, with relatively less liquidity and less farming experience should be given the choice to rent land with an option to purchase. The preference for first renting by most of the surveyed MSFs could indicate that many very highly leveraged MSFs still experience cash flow stress despite the interest rate subsidy. A second policy implication, therefore, is that the current subsidy level, which reduces the effective starting interest rate level to about ten per cent relative to a typical five per cent current return on land, could be increased to promote access to farmland markets. Alternatively, loan terms in the next round of the scheme could be changed to require higher proportions of own equity (lower leverage levels), or to permit the deferral of principal payments, or to permit the purchase of smaller farms by creditworthy, part-time farmers. Another strategy to improve liquidity is to advise growers to limit family drawings in the early years after farmland purchase. / Thesis (M.Sc.Agric.)-University of Natal, Pietermaritzburg, 2003.